RFID-based underground asset management system

Abstract

The present application relates to the technical field of Internet of Things intelligent sensing, and particularly relates to an underground asset management system based on RFID, comprising a scene sensing control module, which is used for monitoring a physical trigger signal of an asset access channel, calculating a concurrency parameter according to a duration of the physical trigger signal, and generating an adaptive acquisition strategy instruction containing a dynamic anti-collision Q value parameter and a multi-stage power polling configuration. In the present application, the scene sensing control module is used for monitoring the physical trigger signal in the channel in real time, quantitatively determining the scene concurrency level, and then adaptively adjusting the anti-collision Q value parameter and the transmission power polling strategy of the reader. This dynamic regulation mechanism effectively solves the problem of missed reading or collision that is prone to occur when the traditional fixed frequency and fixed power mode faces high concurrency asset access in the underground, and significantly improves the radio frequency identification accuracy and data acquisition integrity in the complex electromagnetic environment and variable access density.

Description

Technical Field

[0001] This invention relates to the field of IoT intelligent sensing technology, and in particular to an RFID-based downhole asset management system. Background Technology

[0002] In the field of intelligent construction of coal mines and safety management of underground engineering, the supervision of the entry and exit of mining equipment, tools and materials is a core link in ensuring safe production. It usually relies on ultra-high frequency radio frequency identification technology to deploy reading and writing devices at key checkpoints such as mine entrances or transportation roadways, so as to realize non-contact automatic identification and access information registration of electronic tags attached to mobile assets.

[0003] However, most existing radio frequency acquisition devices use preset fixed transmission power and static anti-collision algorithm parameters, which lack the ability to sense changes in the density of passing targets on site. When faced with high-concurrency scenarios of personnel and assets passing through the mine in a tidal manner or complex electromagnetic environments limited by the multipath effect of the well wall, the fixed radio frequency parameters are difficult to dynamically balance the contradiction between the identification rate and the channel anti-collision efficiency. This can easily lead to tag signal conflicts, data congestion or coverage blind spots, resulting in frequent occurrences of asset omissions, misreads and cross-channel cross-reads, which seriously restricts the real-time performance and accuracy of refined management and control of the entire process of underground assets. Summary of the Invention

[0004] To overcome the above shortcomings, this invention provides an RFID-based downhole asset management system, which aims to improve the problem that most existing radio frequency acquisition devices use preset fixed transmission power and static anti-collision algorithm parameters, lacking the ability to sense changes in the density of on-site traffic targets.

[0005] This invention provides the following technical solution: an RFID-based downhole asset management system, comprising: The scene perception control module is used to monitor the physical trigger signals of the asset access channel, calculate the concurrency parameters based on the duration of the physical trigger signals, and generate an adaptive acquisition strategy instruction that includes dynamic anti-collision Q-value parameters and multi-level power polling configuration. The adaptive radio frequency acquisition module, based on the adaptive acquisition strategy instruction, adjusts the air interface protocol parameters of the reader and performs a step-by-step cyclic scan of the transmit power, and outputs a set of raw radio frequency timing signals including the target electronic tag ID, received signal strength indicator RSSI, phase value and reading timestamp; The signal feature cleaning module, based on the original radio frequency time series signal set, uses a smoothing filtering algorithm to filter outlier data points with numerical abrupt changes in the original radio frequency time series signal set, and calculates the confidence weight of each data point according to the fluctuation variance of the signal strength to generate a weighted intensity feature matrix. The trajectory logic analysis module, based on the weighted strength feature matrix, selects the Received Signal Strength Indicator (RSSI) with confidence weights satisfying a preset threshold for curve fitting analysis, calculates the antenna phase difference based on the phase value, determines the displacement direction based on the temporal change of the antenna phase difference, and generates an initial judgment event for asset entry / exit from the well. Accompanying the correlation verification module, based on the initial judgment event of the asset entry and exit from the well, it retrieves the reference electronic tag ID that coexists with the target electronic tag ID within a preset time window, compares it with the pre-stored correlation object mapping relationship table, filters out asset events that do not match the correlation object, and generates a compliance verification pass instruction; The status association update module, based on the compliance verification pass instruction, indexes the asset database and updates the corresponding location status field to generate an asset management status result.

[0006] Preferably, in the scene perception control module, the step of generating an adaptive acquisition strategy instruction that includes dynamic anti-collision Q-value parameters and multi-level power polling configuration specifically includes the following steps: The effective trigger state duration of the physical trigger signal is obtained, and the effective trigger state duration is compared with a preset congestion determination threshold. When the duration of the effective triggering state is greater than the congestion determination threshold, the current scenario level is determined to be a high-concurrency mode, and a first value is selected as the initial Q-value index; when the duration of the effective triggering state is less than or equal to the congestion determination threshold, the current scenario level is determined to be a low-concurrency mode, and a second value less than the first value is selected as the initial Q-value index. Based on the determined scenario level, the corresponding power cycle scanning parameters are retrieved from the pre-stored power strategy mapping table. The power cycle scanning parameters include the step gradient of the transmit power increasing from the minimum power value to the maximum power value in one scanning cycle and the dwell time of each power level. The initial Q-value exponent is written into the anti-collision parameter field, and the power cyclic scanning parameter is written into the radio frequency control field. The two are combined to generate an adaptive acquisition strategy instruction.

[0007] Preferably, in the adaptive radio frequency acquisition module, the output includes a raw radio frequency timing signal set containing the target electronic tag ID, received signal strength indicator (RSSI), phase value, and read timestamp, specifically including the following steps: Parse the adaptive acquisition strategy instruction, load the dynamic anti-collision Q-value parameter into the anti-collision algorithm register of the reader, and set the initial time slot count value; Start the radio frequency transmission unit, and periodically adjust the transmission gain of the radio frequency front end according to the step gradient and dwell time set in the multi-level power polling configuration, and send continuous wave and inventory command at each level of transmission gain. The backscattered signal of the target electronic tag in response to the inventory command is received, the backscattered signal is demodulated to obtain the electronic tag ID, and the amplitude level and carrier phase difference of the backscattered signal are sampled and converted into Received Signal Strength Indicator (RSSI) and phase value, respectively. The system clock at the demodulation completion time is obtained, and the system clock is used as a read timestamp. It is then encapsulated with the electronic tag ID, the received signal strength indicator RSSI, and the phase value into a single frame signal record, and appended in chronological order to form the original radio frequency timing signal set.

[0008] Preferably, in the signal feature cleaning module, the step of filtering outlier data points with abrupt numerical changes in the original radio frequency time series signal using a smoothing filtering algorithm specifically includes the following steps: The original radio frequency time series signal set is grouped according to the target electronic tag ID to generate several independent time series subsets; For each subset of the time series, a sliding sampling window of length N is established. As new signal records enter, the data within the sliding sampling window is updated according to the first-in-first-out principle. The statistical mean and standard deviation of all received signal strength indicators (RSSI) within the sliding sampling window are continuously calculated. Set a deviation determination coefficient, calculate the difference between the Received Signal Strength Indication (RSSI) and the statistical mean of the current data point, and determine whether the absolute value of the difference is greater than the product of the standard deviation and the deviation determination coefficient; If the judgment result is yes, then the current data point is determined to be an outlier data point with a numerical mutation, and the current data point is removed from the time series subset; if the judgment result is no, then the current data point is determined to be a valid data point and is retained in the time series subset.

[0009] Preferably, in the signal feature cleaning module, the step of calculating the confidence weight of each data point based on the fluctuation variance of the signal intensity and generating a weighted intensity feature matrix specifically includes the following steps: For each valid data point retained by the smoothing filtering algorithm, the statistical variance of the Received Signal Strength Indication (RSSI) within a sliding sampling window corresponding to the valid data point is calculated. Using a preset negative correlation weight mapping function, the statistical variance value is mapped and calculated as a confidence weight; The confidence weights are associated and bound with the corresponding electronic tag ID, received signal strength indicator (RSSI), phase value, and reading timestamp to construct a multi-dimensional vector containing time-domain information and statistical features. All the multidimensional vectors are arranged in the order of their read timestamps to form a weighted intensity feature matrix.

[0010] Preferably, in the trajectory logic analysis module, the step of selecting RSSI data with weight values ​​that meet a preset threshold for curve fitting analysis specifically includes the following steps: Traverse the weighted intensity feature matrix, and for each multidimensional vector contained in the weighted intensity feature matrix, compare the confidence weight in the multidimensional vector with a preset validity threshold, and extract multidimensional vectors greater than the validity threshold to form a fitting sample set; Using the reading timestamp as the independent variable and the Received Signal Strength Indication (RSSI) as the dependent variable, the least squares method is used to perform regression analysis on the fitted sample set to construct a quadratic polynomial trajectory model. Calculate the goodness-of-fit coefficient and quadratic term coefficient of the quadratic polynomial trajectory model. If the goodness-of-fit coefficient is greater than the preset goodness-of-fit threshold and the quadratic term coefficient is negative, then it is determined that the current signal exhibits parabolic distribution characteristics. The peak time corresponding to the peak signal strength is calculated based on the quadratic polynomial trajectory model. The peak time is confirmed as the peak time point when the asset passes through the center of the radio frequency antenna, and the fitted RSSI value corresponding to the peak time is output.

[0011] Preferably, in the accompanying association verification module, the step of comparing the pre-stored association object mapping relationship table and filtering asset events that do not match the association object specifically includes the following steps: Extract the target electronic tag ID from the initial judgment event of the asset entry and exit from the well, and query the associated attribute field corresponding to the target electronic tag ID in the associated object mapping table. The associated attribute field defines whether the asset is an independent movement type or an accompanying binding type. If the associated attribute field is of the independent movement type, the verification is directly considered to have passed; if the associated attribute field is of the accompanying binding type, the set of specified accompanying IDs bound to the target electronic tag ID is obtained. Obtain reference electronic tag IDs that are in the same time window as the target electronic tag ID, and iterate through the reference electronic tag IDs to determine whether any ID in the specified accompanying ID set appears in the reference electronic tag IDs; If the judgment result is yes, the accompanying relationship is confirmed to be established, and a compliance verification pass instruction containing the target electronic tag ID and displacement direction parameters is generated; if the judgment result is no, the accompanying relationship is confirmed to be missing, the current event is marked as abnormal and intercepted, and no compliance verification pass instruction is generated.

[0012] Preferably, in the status association update module, the process of indexing the asset database and updating the corresponding location status field to generate an asset management status result specifically includes the following steps: In response to the compliance verification pass instruction, the corresponding data record item in the asset database is located using the target electronic tag ID as the index key; Read the displacement direction parameter in the compliance verification pass instruction. If the displacement direction parameter is the well entry direction, update the position status field in the data record item to the first position status. If the displacement direction parameter is the well exit direction, update the position status field to the second position status. Obtain the current system time and compare it with the preset maintenance deadline and forced scrapping date in the data record item; If the current system time exceeds the preset maintenance deadline, the asset health status field will be set to maintenance warning status; if the current system time exceeds the mandatory scrapping date, the asset health status field will be set to scrapping and discontinuation status. Summarize the updated location status field and asset health status field, and output the asset management status result in a package.

[0013] The present invention has the following beneficial effects: 1. In this invention, the scene perception control module monitors the physical trigger signals in the channel in real time and quantifies the scene concurrency level, and then adaptively adjusts the anti-collision Q value parameter and transmission power polling strategy of the reader. This dynamic control mechanism effectively solves the problem of missed reading or conflict that is easy to occur when the traditional fixed frequency and fixed power mode faces high concurrency asset passage in the mine, and significantly improves the accuracy of radio frequency identification and the integrity of data acquisition under complex electromagnetic environment and variable passage density.

[0014] 2. In this invention, a sliding sampling window based on statistical principles is constructed using a signal feature cleaning module to remove outliers and noise from numerical abrupt changes. Combined with a quadratic polynomial regression algorithm, trajectory fitting analysis is performed on high-confidence weighted RSSI data. This deep signal processing method can accurately reconstruct the motion characteristics of assets from fluctuating data affected by multipath effects, effectively distinguishing real moving assets from stationary interference sources or wandering misread signals, thereby significantly improving the logical robustness and spatiotemporal positioning accuracy of asset entry and exit events.

[0015] 3. In this invention, the compliance of personnel or vehicle binding is automatically screened by comparing with the pre-stored associated object mapping table, and the asset database is linked to update the location status and health warning status in real time according to the system time. This hardware and software collaborative management mode realizes the leap from simple location tracking to safety and compliance auditing, effectively eliminating the safety hazards of unauthorized carrying, overdue service or uninspected equipment entering the well. Attached Figure Description

[0016] Figure 1This is an architecture diagram of the RFID-based downhole asset management system proposed in this invention. Detailed Implementation

[0017] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] This invention provides an RFID-based downhole asset management system, such as... Figure 1 As shown, it includes: The scene perception control module is used to monitor the physical trigger signals of asset access channels, calculate the concurrency parameters based on the duration of the physical trigger signals, and generate an adaptive acquisition strategy instruction that includes dynamic anti-collision Q-value parameters and multi-level power polling configuration. Furthermore, the scene perception control module generates an adaptive acquisition strategy instruction that includes dynamic anti-collision Q-value parameters and multi-level power polling configuration, specifically including the following steps: The effective trigger state duration of the physical trigger signal is obtained, and the effective trigger state duration is compared with the preset congestion judgment threshold. When the duration of the effective triggering state is greater than the congestion judgment threshold, the current scenario level is determined to be high concurrency mode, and the first value is selected as the initial Q value index; when the duration of the effective triggering state is less than or equal to the congestion judgment threshold, the current scenario level is determined to be low concurrency mode, and the second value less than the first value is selected as the initial Q value index. Based on the determined scenario level, the corresponding power cycle scanning parameters are retrieved from the pre-stored power strategy mapping table. The power cycle scanning parameters include the step gradient of the transmit power from the minimum power value to the maximum power value in a scanning cycle and the dwell time of each power level. The initial Q-value exponent is written into the anti-collision parameter field, and the power cyclic scan parameter is written into the RF control field. These are combined to generate an adaptive acquisition strategy instruction.

[0019] Specifically, the scene perception control module measures the duration of the physical trigger signal at the asset access point to obtain the effective trigger state duration for each trigger event. Let the duration of one effective trigger state be... Its value is the time difference between the start time of the physical trigger signal switching from an invalid state to an effective state and the end time of switching back from an effective state to an invalid state, in seconds or milliseconds. The preset congestion judgment threshold is denoted as... This threshold is a fixed time constant set by the system during the deployment phase based on the physical size of the channel, the average transit time of a single asset, and historical traffic statistics.

[0020] When satisfied When the scene perception control module determines that multiple assets pass through the channel consecutively or overlappingly within the trigger period, the current channel scene level is determined to be high-concurrency mode; when the conditions are met... When this occurs, it is determined to be in low-concurrency mode. The input parameter is derived from the time measurement result of the physical trigger signal, and its comparison output result is the scene level determination information, which is used to drive the adaptive configuration of subsequent radio frequency parameters.

[0021] After determining the scene level, the scene perception and control module generates the initial Q-value exponent corresponding to the anti-collision algorithm, and records the initial Q-value exponent as . ,in Let be a non-negative integer. When the scene level is high concurrency mode, let... When the scene level is low concurrency mode, make ,and ,in, This represents a larger Q-value exponent preset for high-concurrency scenarios, used to increase the number of time slots in the anti-collision algorithm; This represents a smaller Q-value index preset for low-concurrency scenarios, used to reduce idle time slots and shorten inventory counting time.

[0022] From the initial Q value exponent The number of time slots in the anti-collision inventory polling is calculated. The calculation formula is as follows: ; in, This indicates the total number of time slots available in a single collision avoidance inventory cycle, used to limit the range of time slots that electronic tags can randomly select for response; The exponential parameter written to the anti-collision algorithm register of the reader directly affects the tag response collision probability and inventory efficiency.

[0023] Meanwhile, the scene perception control module retrieves power cyclic scanning parameters from a pre-stored power strategy mapping table based on the scene level, setting the minimum transmit power to be... Maximum transmission power is The power step gradient is The single-stage power residence time is ,in, and These represent the minimum and maximum transmit power values ​​that the RFID reader is allowed to configure under the current hardware conditions, respectively, in decibels and milliwatts; This represents the incremental value between the transmit power of two adjacent stages; This indicates the length of time that the radio frequency front end remains constant at a certain transmit power.

[0024] Within one power scan cycle, the first Level transmit power The calculation method is as follows: ; in, For power level indices, use zero-based integers, and satisfy the following conditions: Each power level The duration is It is used to collect and inventory electronic tags in the channel under different radio frequency coverage and signal strength conditions.

[0025] After completing the above parameter calculations, the initial Q-value exponent As an anti-collision parameter, it is written into the anti-collision parameter field; power cycle scan parameter. , , and As radio frequency control parameters, they are written into the radio frequency control field and jointly encapsulated to generate adaptive acquisition strategy instructions, which are used to control the working mode of the subsequent adaptive radio frequency acquisition module.

[0026] Through the above steps, the scene perception control module realizes a closed-loop mapping from channel physical trigger signal to concurrent scene quantization judgment, and then to anti-collision and power control parameter generation, enabling the radio frequency acquisition strategy to dynamically match the actual asset traffic density, and providing clear and reproducible pre-control conditions for the subsequent stable generation of the original radio frequency timing signal set.

[0027] The adaptive radio frequency acquisition module, based on the adaptive acquisition strategy command, adjusts the air interface protocol parameters of the reader and performs a step-by-step cyclic scan of the transmit power, and outputs a set of raw radio frequency timing signals including the target electronic tag ID, received signal strength indicator RSSI, phase value and reading timestamp; Furthermore, the adaptive RF acquisition module outputs a raw RF timing signal set containing the target electronic tag ID, Received Signal Strength Indicator (RSSI), phase value, and read timestamp, specifically including the following steps: Parse the adaptive acquisition strategy instruction, load the dynamic anti-collision Q value parameter into the anti-collision algorithm register of the reader, and set the initial time slot count value; Start the RF transmitting unit and periodically adjust the transmit gain of the RF front end according to the step gradient and dwell time set in the multi-level power polling configuration. At each level of transmit gain, send continuous wave and inventory command. The backscattered signal of the target electronic tag in response to the inventory command is received, the backscattered signal is demodulated to obtain the electronic tag ID, and the amplitude level and carrier phase difference of the backscattered signal are sampled and converted into Received Signal Strength Indicator (RSSI) and phase value, respectively. The system clock at the demodulation completion time is obtained. The system clock is used as a read timestamp and encapsulated with the electronic tag ID, received signal strength indicator (RSSI), and phase value into a single frame signal record. These are then appended in chronological order to form the original radio frequency timing signal set.

[0028] Specifically, the adaptive RF acquisition module is connected to the aforementioned scene perception control module via a control bus or high-speed serial interface. It is used to receive and parse the generated adaptive acquisition strategy instructions. The dynamic anti-collision Q-value parameters and multi-level power polling configuration parameters contained in the adaptive acquisition strategy instructions are encapsulated in the form of register configuration fields. After receiving the instruction, the adaptive RF acquisition module first parses the instruction content, extracts the dynamic anti-collision Q-value parameters and loads them into the anti-collision algorithm register corresponding to the reader's air interface protocol. At the same time, it clears the time slot counter or sets it to the initial count value to mark the start state of an anti-collision inventory polling.

[0029] After completing the anti-collision parameter configuration, the adaptive RF acquisition module starts the RF transmission unit. Based on the minimum transmission power, maximum transmission power, power step gradient, and single-level power dwell time set in the multi-level power polling configuration, it performs a step-by-step cyclic scan of the RF front-end's transmission gain. At each power level, the RF transmission unit first outputs a continuous wave signal to activate the electronic tags within the current coverage area. Then, it sends an inventory command according to the configured anti-collision parameters, causing the target electronic tag to return a backscatter signal in the corresponding time slot. The adjustment and dwell time of the transmission power are completed by the local timing control logic, ensuring that at least one complete anti-collision inventory poll can be completed at each power level.

[0030] During the receiving phase, the adaptive RF acquisition module performs down-conversion and analog-to-digital conversion on the backscattered signal returned by the electronic tag through the RF receiving unit to obtain the baseband signal. The baseband signal is then demodulated to recover the unique identification information of the electronic tag, and the electronic tag identifier is recorded as follows: subscript Indicates the first Upon successful read, the amplitude component of the baseband signal is sampled to obtain the received signal amplitude value. The phase components of the carrier are estimated to obtain the phase difference value. .

[0031] The Received Signal Strength Indicator (RSSI) is calculated from the received signal amplitude value, and the calculation relationship is as follows: ; in, Indicates the first The received signal strength indication value of the backscattered signal of the electronic tag read once, in decibels; This represents the amplitude sample value of the corresponding backscattered baseband signal. This amplitude value originates from the analog-to-digital converter output of the RF receiving link, and the phase value... It represents the phase difference of the received signal relative to the local reference carrier under the current transmit power and time slot conditions, and is used to characterize the electromagnetic propagation characteristics between the tag and the antenna.

[0032] After acquiring the electronic tag ID, RSSI, and phase value, the adaptive RF acquisition module reads the current system clock count value and records it as... As the first The timestamp of the tag reading event is used. The system clock is a high-precision timing module inside the reader, and its time base is synchronized with the transmission power switching and inventory command sending process. Subsequently, the adaptive RF acquisition module will... , , and It is encapsulated as a single-frame signal record and stored or output in ascending order of timestamp.

[0033] By repeating the above process under multiple transmit power levels and multiple anti-collision time slots, the adaptive RF acquisition module continuously generates an original RF timing signal set consisting of multiple single-frame signal records. This signal set completely preserves the response characteristics of the electronic tag under different times and different RF conditions.

[0034] Through the above implementation steps, the adaptive RF acquisition module can dynamically adjust the air interface parameters and perform stepped power scanning under the constraints of the aforementioned adaptive acquisition strategy instructions, stably outputting a high-time-resolution original RF timing signal set that matches the channel scenario, providing reliable input data for subsequent identification and analysis based on timing features.

[0035] The signal feature cleaning module, based on the original radio frequency time series signal set, uses a smoothing filtering algorithm to filter out outlier data points with numerical abrupt changes in the original radio frequency time series signal set, and calculates the confidence weight of each data point according to the fluctuation variance of the signal strength to generate a weighted intensity feature matrix. Furthermore, in the signal feature cleaning module, a smoothing filtering algorithm is used to filter out outlier data points with abrupt numerical changes in the original RF time series signal set. This specifically includes the following steps: The original radio frequency time series signal set is grouped according to the target electronic tag ID to generate several independent time series subsets; For each subset of the time series, a sliding sampling window of length N is established. As new signal records enter, the data in the sliding sampling window is updated according to the first-in-first-out principle. Continuously calculate the statistical mean and standard deviation of all received signal strength indicators (RSSI) within the sliding sampling window; Set a deviation determination coefficient, calculate the difference between the received signal strength index (RSSI) of the current data point and the statistical mean, and determine whether the absolute value of the difference is greater than the product of the standard deviation and the deviation determination coefficient; If the judgment result is yes, the current data point is determined to be an outlier data point with a numerical mutation, and the current data point is removed from the time series subset; if the judgment result is no, the current data point is determined to be a valid data point and is retained in the time series subset.

[0036] Furthermore, in the signal feature cleaning module, the confidence weight of each data point is calculated based on the fluctuation variance of the signal strength to generate a weighted intensity feature matrix, specifically including the following steps: For each valid data point retained by the smoothing filtering algorithm, the statistical variance of the Received Signal Strength Indication (RSSI) within a sliding sampling window corresponding to the valid data point is calculated. Using a pre-defined negative correlation weight mapping function, the statistical variance value is mapped and calculated as a confidence weight; The confidence weights are associated and bound with the corresponding electronic tag ID, received signal strength indicator (RSSI), phase value, and reading timestamp to construct a multi-dimensional vector containing time-domain information and statistical features. All multidimensional vectors are arranged in the order of reading timestamps to form a weighted intensity feature matrix.

[0037] Specifically, the signal feature cleaning module and the adaptive RF acquisition module are connected via a data bus. The module receives the original RF timing signal set output by the module as input data. The original RF timing signal set contains several single-frame signal records arranged in chronological order. Each frame record carries an electronic tag ID, a received signal strength indicator (RSSI), a phase value, and a reading timestamp. The signal feature cleaning module first reads the original RF timing signal set and divides the data into several independent time series subsets using the electronic tag ID as the key field, ensuring that subsequent filtering processing is performed on the timing characteristics of a single physical target.

[0038] For each subset of time series, construct a length of The sliding sampling window, as new signal is recorded According to timestamp The sequences are entered sequentially, and the sliding sampling window moves forward along the time axis, always including the time before the current moment. For each historical data point within the window, statistical methods are used to calculate the local mean and standard deviation. Let the nth... The data set within each sliding window is ,in Indicates the first in the window The RSSI values ​​of each sampling point, the statistical mean of this window. with standard deviation The calculation formula is as follows: ; ; in, The received signal strength indication value is derived from the original radio frequency timing signal set, and the unit is dBm; This is a preset window length integer value, typically between 5 and 20, used to balance real-time computation with smoothing effects; This indicates the central trend of signal strength within the current local time range; It indicates the degree of signal dispersion within the current local time range.

[0039] After calculating the statistical parameters, set the deviation determination coefficient. For the current data point to be determined The outlier detection logic is as follows: ; in, The RSSI sample value currently entering the decision process; A preset positive real number (e.g., 2 or 3) is used to define the outlier boundary outside the confidence interval of the normal distribution. If the above inequality holds, a determination is made. Outlier mutations caused by environmental multipath effects or hardware noise are physically removed from the time series subset. If the condition is not met, the data point is retained as valid data.

[0040] For each valid data point retained after smoothing filtering, the signal feature cleaning module further calculates its confidence weight and, for the sliding window containing the valid data point, calculates the statistical variance of the RSSI values ​​within that window. The calculation formula is as follows: ; in, It directly reflects the intensity of radio frequency signal fluctuations in the vicinity at that moment. A smaller value indicates a more stable signal. Using a preset negative correlation weighting mapping function, the signal is... Convert to normalized confidence weights The calculation formula is: ; in, The sensitivity coefficient for adjusting the weights is set to a positive real number. The confidence weights for the output have a range of values. This formula ensures that data points with larger signal fluctuation variance have lower confidence weights.

[0041] Finally, the signal feature cleaning module will calculate the confidence weights. Compared with the electronic tag ID and RSSI value in the original data record Phase value and read timestamp Perform association binding to construct multidimensional feature vectors All processed multidimensional feature vectors are sorted by timestamp. The order of the elements is rearranged to form a weighted intensity feature matrix, which is then output.

[0042] Through the above steps, the signal feature cleaning module can adaptively remove impulse noise interference based on the statistical characteristics of the original radio frequency data, and provide weighted data carrying quality assessment information for subsequent positioning analysis, thereby significantly improving the robustness and accuracy of asset trajectory fitting in complex downhole environments.

[0043] The trajectory logic analysis module, based on the weighted strength feature matrix, selects the Received Signal Strength Indicator (RSSI) with confidence weights satisfying a preset threshold for curve fitting analysis, calculates the antenna phase difference based on the phase value, determines the displacement direction based on the temporal change of the antenna phase difference, and generates an initial judgment event for asset entry / exit from the well. Furthermore, in the trajectory logic analysis module, RSSI data with weight values ​​that meet preset thresholds are selected for curve fitting analysis, specifically including the following steps: Traverse the weighted intensity feature matrix, and for each multi-dimensional vector contained in the weighted intensity feature matrix, compare the confidence weight in the multi-dimensional vector with the preset validity threshold, and extract the multi-dimensional vectors that are greater than the validity threshold to form the fitting sample set; Using the reading timestamp as the independent variable and the Received Signal Strength Indicator (RSSI) as the dependent variable, the least squares method was used to perform regression analysis on the fitted sample set to construct a quadratic polynomial trajectory model. Calculate the goodness-of-fit coefficient and quadratic term coefficient of the quadratic polynomial trajectory model. If the goodness-of-fit coefficient is greater than the preset goodness-of-fit threshold and the quadratic term coefficient is negative, then the current signal is determined to exhibit parabolic distribution characteristics. The peak time corresponding to the peak signal strength is calculated based on the quadratic polynomial trajectory model. The peak time is identified as the peak time point when the asset passes through the center of the radio frequency antenna, and the fitted RSSI value corresponding to the peak time is output.

[0044] Specifically, the trajectory logic analysis module obtains the weighted intensity feature matrix as input data from the signal feature cleaning module. The weighted intensity feature matrix consists of several multi-dimensional feature vectors arranged in chronological order. Each multi-dimensional feature vector contains the electronic tag ID, reading timestamp, Received Signal Strength Indicator (RSSI), phase value, and confidence weight. The module first performs weight-based filtering on the data in the matrix, traversing each multi-dimensional feature vector, extracting its confidence weight, and comparing it with a preset validity threshold. The confidence weight is set to... The preset validity threshold is ,in This is a dimensionless constant set according to the system's noise tolerance, with a value ranging from 0 to 1, when the following conditions are met. When the data point corresponding to the vector is determined to have statistical value for participating in trajectory analysis, its reading timestamp and Received Signal Strength Indicator (RSSI) are extracted and stored in the fitted sample set.

[0045] After constructing the fitted sample set, the trajectory logic analysis module establishes a quadratic regression analysis model based on the least squares principle. Let the fitted sample set contain a total of... The number of valid data points, the first The read timestamp of each data point is recorded as follows: The corresponding received signal strength indication value is recorded as The constructed quadratic polynomial trajectory model expression is as follows: ; in, Indicates time The predicted received signal strength at any given time, in dBm; The coefficient of the quadratic term represents the curvature and direction of the opening of the signal strength change curve, which is physically related to the speed and distance of the asset passing through the antenna beam. The coefficient of the linear term represents the slope of the curve; is a constant term that characterizes the intercept level of the curve.

[0046] To solve for the optimal model coefficients, the module constructs a design matrix. With observation vector And solve the coefficient vector using the normal equation system. The calculation formula is as follows: ; In the above formula, the coefficient vector Observation vector It consists of RSSI measurements from the sample set, and a matrix is ​​designed. for The Vandermonde matrix, its th row element is This step directly obtains the coefficient solution that minimizes the sum of squared residuals through matrix operations.

[0047] After obtaining the model coefficients, the module calculates the goodness-of-fit coefficients. And combined with the coefficient of the quadratic term Perform feature validation and goodness-of-fit coefficient calculation. The calculation formula is: ; in, To timestamp Substitute the predicted RSSI value obtained from the model calculation. The module will calculate the arithmetic mean of all observed RSSI values ​​in the sample set. With the preset goodness-of-fit threshold Compare and examine The symbol, if and only if and At that time, it was determined that the current signal sequence exhibited a significant inverted parabolic distribution characteristic, confirming that the electronic tag had undergone a complete cross-antenna beam behavior.

[0048] Finally, based on the validated quadratic polynomial trajectory model, the time point corresponding to the peak signal strength is calculated, and the peak time point is determined. Determined by the vertex formula of a quadratic function: ; in, The module substitutes the moment when the physical location of the asset is closest to the geometric center of the RF antenna into the original model to calculate the fitted peak intensity. and will , The corresponding electronic tag ID is used to encapsulate and generate the initial judgment event output for asset entry and exit from the well.

[0049] Through the above steps, the trajectory logic analysis module uses a weighted regression algorithm to reconstruct a continuous asset movement trajectory from discrete and fluctuating radio frequency measurement data. It effectively distinguishes between real passing events and stationary interference signals by utilizing the geometric features of quadratic curves, and accurately obtains the key time nodes of asset passing, providing a high-precision time reference for subsequent verification of the accompanying relationship.

[0050] The accompanying correlation verification module, based on the initial judgment event of asset entry and exit from the well, retrieves reference electronic tag IDs that coexist with the target electronic tag ID within a preset time window, compares them with the pre-stored correlation object mapping relationship table, filters out asset events that do not match the correlation object, and generates a compliance verification pass instruction; Furthermore, in the associated verification module, the pre-stored associated object mapping table is compared to filter asset events that do not match associated objects. This specifically includes the following steps: Extract the target electronic tag ID from the initial judgment event of asset entry and exit from the well, and query the associated attribute field corresponding to the target electronic tag ID in the associated object mapping table. The associated attribute field defines whether the asset is an independent movement type or an accompanying binding type. If the associated attribute field is of the independent movement type, the verification is directly considered to have passed; if the associated attribute field is of the accompanying binding type, the set of specified accompanying IDs bound to the target electronic tag ID is obtained. Obtain reference electronic tag IDs that are in the same time window as the target electronic tag ID, and iterate through the reference electronic tag IDs to determine whether any ID in the specified set of accompanying IDs appears in the reference electronic tag IDs; If the judgment result is yes, the accompanying relationship is confirmed to be established, and a compliance verification pass instruction containing the target electronic tag ID and displacement direction parameters is generated; if the judgment result is no, the accompanying relationship is confirmed to be missing, the current event is marked as abnormal and intercepted, and no compliance verification pass instruction is generated.

[0051] Specifically, the accompanying verification module, acting as middleware connecting the physical perception layer and the business logic layer, receives the initial asset entry / exit event output by the trajectory logic analysis module through the system's internal message queue. The data payload of the initial asset entry / exit event is parsed to extract the target asset's electronic tag ID confirmed to be passing through the radio frequency area, as well as the peak time point at which the asset passes the antenna center. Let the target asset's electronic tag ID be... The peak time point is Module utilization As a unique index key, it accesses the mapping table of related objects stored in the local cache or backend server. This table predefines the control category of each downhole asset, and the fields include asset attribute identifiers and binding association lists.

[0052] The module first reads the asset attribute identifier returned by the query, and executes branch logic control based on the identifier content. When the asset attribute identifier is of the independent movement type, it indicates that the asset is authorized to enter and exit the passage independently. The module immediately determines that the current passage operation is compliant and directly enters the instruction generation process. When the asset attribute identifier is of the accompanying binding type, it indicates that the asset must move together with a specific personnel tag or vehicle tag. The module then extracts the electronic tag IDs of all legal accompanying objects that the asset is allowed to be bound to from the binding association list field, and forms these IDs into a specified accompanying ID set, denoted as ______. The set contains at least one pre-defined valid accompanying tag ID.

[0053] To verify whether there are legitimate accompanying objects during actual passage, the module uses peak time points. Construct an accompanying decision time window for the time base, and set the tolerance radius of the time window as . The effective time retrieval interval is ,in This is a constant calculated based on the downhole travel speed and the length of the radio frequency coverage area, in seconds. The module retrieval system records all data except... within the time retrieval interval. Other than the electronic tag IDs, these IDs are deduplicated and combined to form a reference electronic tag ID set, denoted as . , This represents the set of other tags that actually coexist in physical space around the time the target asset passes through the antenna.

[0054] After constructing the two sets, the module performs a set intersection operation to determine whether the adjoint relationship holds, and defines an adjoint verification result variable. Its calculation logic is determined by the following set operation formulas: ; in, This represents the intersection operation of sets; Indicates the empty set; This indicates that the association relationship verification has passed; The formula means that if the set of environmental reference tags collected within the current time window contains any one or more IDs from the specified set of accompanying IDs, then the physical accompanying relationship is considered to be established.

[0055] Based on the calculated associated verification result variables The module performs the final instruction generation operation when Upon confirming the compliance of the current asset carrying behavior, the module extracts the displacement direction parameters from the original event, generates a compliance verification pass instruction containing the target electronic tag ID and displacement direction parameters, and sends it to the next-level status association update module; when If the current asset is confirmed to be in a state of unauthorized independent movement or illegal carrying, the module will intercept the event, not generate a compliance verification pass instruction, and trigger an abnormal alarm log recording.

[0056] Through the above steps, the accompanying correlation verification module superimposes strict business correlation logic on the trajectory perception at the pure physical level. By using a spatiotemporal correlation algorithm based on set theory, it accurately filters out unauthorized asset movement events, ensuring that important underground equipment must comply with the management specifications of personnel-capital binding or vehicle-capital binding, thereby achieving closed-loop verification of asset security control.

[0057] The status association update module, based on the compliance verification pass instruction, indexes the asset database and updates the corresponding location status field to generate asset management status results.

[0058] Furthermore, in the status association update module, the asset database is indexed and the corresponding location status field is updated to generate the asset management status result, specifically including the following steps: Upon receiving the compliance verification pass command, the corresponding data record item in the asset database is located using the target electronic tag ID as the index key; Read the displacement direction parameter in the compliance verification instruction. If the displacement direction parameter is the well entry direction, update the position status field in the data record item to the first position status. If the displacement direction parameter is the well exit direction, update the position status field to the second position status. Obtain the current system time and compare it with the preset maintenance deadline and forced scrapping date in the data record items respectively; If the current system time exceeds the preset maintenance deadline, the asset health status field will be set to maintenance warning status; if the current system time exceeds the mandatory scrapping date, the asset health status field will be set to scrapping and discontinuation status. Summarize the updated location status field and asset health status field, and output the asset management status result in a package.

[0059] Specifically, the status association update module is deployed in the data processing layer of the backend server. It receives compliance verification pass instructions from the association verification module via an internal bus or network interface. The data payload of this instruction includes the target electronic tag ID, which has undergone multiple verifications, and displacement direction parameters representing the asset's movement direction. The module first parses the instruction, extracts the target electronic tag ID, and denotes it as... ,use As a unique primary key index, it is used to retrieve the corresponding asset data record in the asset database. The asset database is a non-volatile storage system that stores information about the entire lifecycle of underground assets. Each record... Includes current location status field Asset health status field Preset maintenance deadline and the mandatory scrapping date .

[0060] After locking the target record, the module performs an update operation on the position status field based on the displacement direction parameter. Let the displacement direction parameter in the instruction be... This parameter originates from the analytical results of the antenna phase difference timing changes in the trajectory logic analysis module. The state constant of the downhole working area is defined as follows: The ground / warehouse area state constant is Location status field The update logic is defined by the following state transition function: ; in, A feature indicating that an asset moves from outside the wellhead radio frequency coverage area inward and crosses the antenna beam; This feature indicates that an asset has moved from inside the wellhead radio frequency coverage area to the outside. The formula establishes a mapping between physical spatial displacement and digital spatial status, ensuring that the location information in the database is strictly synchronized with the actual physical location of the asset.

[0061] Subsequently, the module calls the high-precision system clock to obtain the current processing time, denoted as . It also calculates and updates the lifecycle health of assets in real time, and reads the maintenance deadline from the record items. With mandatory scrapping date These two time parameters are based on the asset's manufacturing date, design life, and a pre-set timestamp according to industry safety standards. The asset health status field... The update determination logic is as follows: ; in, This indicates a forced scrapping and decommissioning status. When the system time exceeds the scrapping deadline, the system will forcibly lock this status to prevent the asset from being put back into the well. This indicates a maintenance warning status, suggesting that the asset has exceeded its maintenance period and requires inspection. This indicates that the original state of health will remain unchanged. The server's UTC time at the time the current update operation was performed.

[0062] After completing the above calculations and field modifications, the module commits the database transaction, updating the location status field. Asset Health Status Field Persistent writing to the storage medium, at the same time, the module will Updated and Encapsulated as an asset management status result vector And release it to the outside world.

[0063] Through the above steps, the status association update module realizes the automated flow from radio frequency sensing physical events to enterprise-level management status. Utilizing a strict time logic judgment formula, it automatically triggers full life cycle health management while completing asset location tracking, effectively preventing overdue or unmaintained assets from being illegally put into the well, and realizing a closed-loop safety mechanism for downhole asset management.

[0064] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. An RFID-based downhole asset management system, characterized in that, include: The scene perception control module is used to monitor the physical trigger signals of the asset access channel, calculate the concurrency parameters based on the duration of the physical trigger signals, and generate an adaptive acquisition strategy instruction that includes dynamic anti-collision Q-value parameters and multi-level power polling configuration. The adaptive radio frequency acquisition module, based on the adaptive acquisition strategy instruction, adjusts the air interface protocol parameters of the reader and performs a step-by-step cyclic scan of the transmit power, and outputs a set of raw radio frequency timing signals including the target electronic tag ID, received signal strength indicator RSSI, phase value and reading timestamp; The signal feature cleaning module, based on the original radio frequency time series signal set, uses a smoothing filtering algorithm to filter outlier data points with numerical abrupt changes in the original radio frequency time series signal set, and calculates the confidence weight of each data point according to the fluctuation variance of the signal strength to generate a weighted intensity feature matrix. The trajectory logic analysis module, based on the weighted strength feature matrix, selects the Received Signal Strength Indicator (RSSI) with confidence weights satisfying a preset threshold for curve fitting analysis, calculates the antenna phase difference based on the phase value, determines the displacement direction based on the temporal change of the antenna phase difference, and generates an initial judgment event for asset entry / exit from the well. Accompanying the correlation verification module, based on the initial judgment event of the asset entry and exit from the well, it retrieves the reference electronic tag ID that coexists with the target electronic tag ID within a preset time window, compares it with the pre-stored correlation object mapping relationship table, filters out asset events that do not match the correlation object, and generates a compliance verification pass instruction; The status association update module, based on the compliance verification pass instruction, indexes the asset database and updates the corresponding location status field to generate an asset management status result.

2. The RFID-based downhole asset management system according to claim 1, characterized in that, In the scene perception control module, the generation of adaptive acquisition strategy instructions, which includes dynamic anti-collision Q-value parameters and multi-level power polling configuration, specifically includes the following steps: The effective trigger state duration of the physical trigger signal is obtained, and the effective trigger state duration is compared with a preset congestion determination threshold. When the duration of the effective triggering state is greater than the congestion determination threshold, the current scenario level is determined to be a high-concurrency mode, and a first value is selected as the initial Q-value index; when the duration of the effective triggering state is less than or equal to the congestion determination threshold, the current scenario level is determined to be a low-concurrency mode, and a second value less than the first value is selected as the initial Q-value index. Based on the determined scenario level, the corresponding power cycle scanning parameters are retrieved from the pre-stored power strategy mapping table. The power cycle scanning parameters include the step gradient of the transmit power increasing from the minimum power value to the maximum power value in one scanning cycle and the dwell time of each power level. The initial Q-value exponent is written into the anti-collision parameter field, and the power cyclic scanning parameter is written into the radio frequency control field. The two are combined to generate an adaptive acquisition strategy instruction.

3. The RFID-based downhole asset management system according to claim 1, characterized in that, In the adaptive radio frequency acquisition module, the output includes a raw radio frequency timing signal set containing the target electronic tag ID, received signal strength indicator (RSSI), phase value, and read timestamp, specifically including the following steps: Parse the adaptive acquisition strategy instruction, load the dynamic anti-collision Q-value parameter into the anti-collision algorithm register of the reader, and set the initial time slot count value; Start the radio frequency transmission unit, and periodically adjust the transmission gain of the radio frequency front end according to the step gradient and dwell time set in the multi-level power polling configuration, and send continuous wave and inventory command at each level of transmission gain. The backscattered signal of the target electronic tag in response to the inventory command is received, the backscattered signal is demodulated to obtain the electronic tag ID, and the amplitude level and carrier phase difference of the backscattered signal are sampled and converted into Received Signal Strength Indicator (RSSI) and phase value, respectively. The system clock at the demodulation completion time is obtained, and the system clock is used as a read timestamp. It is then encapsulated with the electronic tag ID, the received signal strength indicator RSSI, and the phase value into a single frame signal record, and appended in chronological order to form the original radio frequency timing signal set.

4. The RFID-based downhole asset management system according to claim 1, characterized in that, In the signal feature cleaning module, the step of filtering outlier data points with abrupt numerical changes in the original radio frequency time series signal using a smoothing filtering algorithm specifically includes the following steps: The original radio frequency time series signal set is grouped according to the target electronic tag ID to generate several independent time series subsets; For each subset of the time series, a sliding sampling window of length N is established. As new signal records enter, the data within the sliding sampling window is updated according to the first-in-first-out principle. The statistical mean and standard deviation of all received signal strength indicators (RSSI) within the sliding sampling window are continuously calculated. Set a deviation determination coefficient, calculate the difference between the Received Signal Strength Indication (RSSI) and the statistical mean of the current data point, and determine whether the absolute value of the difference is greater than the product of the standard deviation and the deviation determination coefficient; If the judgment result is yes, then the current data point is determined to be an outlier data point with a numerical mutation, and the current data point is removed from the time series subset; if the judgment result is no, then the current data point is determined to be a valid data point and is retained in the time series subset.

5. The RFID-based downhole asset management system according to claim 1, characterized in that, In the signal feature cleaning module, the step of calculating the confidence weight of each data point based on the fluctuation variance of the signal intensity and generating a weighted intensity feature matrix specifically includes the following steps: For each valid data point retained by the smoothing filtering algorithm, the statistical variance of the Received Signal Strength Indication (RSSI) within a sliding sampling window corresponding to the valid data point is calculated. Using a preset negative correlation weight mapping function, the statistical variance value is mapped and calculated as a confidence weight; The confidence weights are associated and bound with the corresponding electronic tag ID, received signal strength indicator (RSSI), phase value, and reading timestamp to construct a multi-dimensional vector containing time-domain information and statistical features. All the multidimensional vectors are arranged in the order of their read timestamps to form a weighted intensity feature matrix.

6. The RFID-based downhole asset management system according to claim 1, characterized in that, In the trajectory logic analysis module, the selection of RSSI data with weight values ​​that meet a preset threshold for curve fitting analysis specifically includes the following steps: Traverse the weighted intensity feature matrix, and for each multidimensional vector contained in the weighted intensity feature matrix, compare the confidence weight in the multidimensional vector with a preset validity threshold, and extract multidimensional vectors greater than the validity threshold to form a fitting sample set; Using the reading timestamp as the independent variable and the Received Signal Strength Indication (RSSI) as the dependent variable, the least squares method is used to perform regression analysis on the fitted sample set to construct a quadratic polynomial trajectory model. Calculate the goodness-of-fit coefficient and quadratic term coefficient of the quadratic polynomial trajectory model. If the goodness-of-fit coefficient is greater than the preset goodness-of-fit threshold and the quadratic term coefficient is negative, then it is determined that the current signal exhibits parabolic distribution characteristics. The peak time corresponding to the peak signal strength is calculated based on the quadratic polynomial trajectory model. The peak time is confirmed as the peak time point when the asset passes through the center of the radio frequency antenna, and the fitted RSSI value corresponding to the peak time is output.

7. The RFID-based downhole asset management system according to claim 1, characterized in that, In the accompanying association verification module, the comparison with the pre-stored association object mapping table to filter asset events that do not match the associated objects specifically includes the following steps: Extract the target electronic tag ID from the initial judgment event of the asset entry and exit from the well, and query the associated attribute field corresponding to the target electronic tag ID in the associated object mapping table. The associated attribute field defines whether the asset is an independent movement type or an accompanying binding type. If the associated attribute field is of the independent movement type, the verification is directly considered to have passed; if the associated attribute field is of the accompanying binding type, the set of specified accompanying IDs bound to the target electronic tag ID is obtained. Obtain reference electronic tag IDs that are in the same time window as the target electronic tag ID, and iterate through the reference electronic tag IDs to determine whether any ID in the specified accompanying ID set appears in the reference electronic tag IDs; If the judgment result is yes, the accompanying relationship is confirmed to be established, and a compliance verification pass instruction containing the target electronic tag ID and displacement direction parameters is generated; if the judgment result is no, the accompanying relationship is confirmed to be missing, the current event is marked as abnormal and intercepted, and no compliance verification pass instruction is generated.

8. The RFID-based downhole asset management system according to claim 1, characterized in that, In the status association update module, the process of indexing the asset database and updating the corresponding location status field to generate an asset management status result includes the following steps: In response to the compliance verification pass instruction, the corresponding data record item in the asset database is located using the target electronic tag ID as the index key; Read the displacement direction parameter in the compliance verification pass instruction. If the displacement direction parameter is the well entry direction, update the position status field in the data record item to the first position status. If the displacement direction parameter is the well exit direction, update the position status field to the second position status. Obtain the current system time and compare it with the preset maintenance deadline and forced scrapping date in the data record item; If the current system time exceeds the preset maintenance deadline, the asset health status field is set to maintenance warning status; if the current system time exceeds the mandatory scrapping date, the asset health status field is set to scrapping and discontinuation status. Summarize the updated location status field and asset health status field, and output the asset management status result in a package.

Images